different solubilities? Was solubility proportional to density and complexity? At this stage Dalton clearly thought solubilities were a function of the sizes of particles3:
I am nearly persuaded that the circumstance depends upon the weight and number of the ultimate particles of the several gases: those whose particles are lightest and single being least absorbable, and others more according as they increase in weight and complexity.
One can see how this line of reasoning would lead automatically to ‘an inquiry into the relative weights of the ultimate particles of bodies … a subject as far as I know, entirely new’. It is important to realize, however, that Dalton really needed to know the weights of particles only because he wanted an estimate of their sizes from the simple relationship, density = weight/volume.
As we have seen, at the end of 1803 Dalton estimated the weights of gas atoms using known chemical analyses and the rule of simplicity. From these he derived a number of atomic volumes and radii, but was unable to find any simple or regular correlation with solubility. Even so, as late as 1810 in the New System he continued to record atomic sizes alongside atomic weights.
It was evidently not until 1804 that Dalton realized that relative atomic weights were a useful explanation of the law of constant composition and that the simple rules of chemical synthesis from which he had derived them explained and predicted that, when elements combined to form more than one compound, the weights of one element that combined with a fixed weight of the other were bound to be small whole numbers. For example if
then the weights of A combined with weight B are in the simple ratio of 1:2. Dalton drew attention to the fact that this was the ratio of hydrogen to carbon in methane (CH4) and ethane (C2H4), and that in the difficult and complicated cases of the oxides of nitrogen the ratio of oxygen to nitrogen was 1 : 2 and 1:3.
There were a large number of cases of this ‘law of multiple proportions’ that had been reported in the literature as a result of the dispute between Proust and Berthollet. When Berthollet accompanied Napoleon’s expedition to Egypt in 1798, he was surprised to find huge deposits of soda by the shores of salt lakes. Mineralogical analysis showed that the soda arose from a reaction between salt and limestone in the lake bottom, in complete contradiction to the usual laboratory reaction in which soda [sodium carbonate] and calcium chloride reacted to form salt and limestone [calcium carbonate]. He concluded that the enormous concentration of salt in the lakes had forced the reversal of the usual reaction. In other words, the action of mass (concentration) could overcome the usual play of elective affinities between substances. In modern terms, Berthollet had stumbled upon an equilibrium reaction:
CaCl2 + Na2CO3
It was this awareness of the role of mass in reactions that caused Berthollet during the next few years to challenge the usual implicit presumption of chemists that substances combined together in fixed proportions, or that constant saturation proportions always characterized chemical union. Instead, Berthollet proposed that compounds combined together in variable and indefinite proportions, and he pointed to solutions and alloys, and what would today be defined as mixtures, as empirical evidence for his claim.
This radical reconceptualization of composition was immediately challenged by a fellow Frenchman, Joseph-Louis Proust (1754–1826), who worked as an analyst in Spain. In a long series of meticulous analyses and friendly challenges to Berthollet, Proust argued that there was overwhelming evidence that regular compounds were formed from their constituents in fixed and definite proportions. There might well be more than one compound of the same two substances, but their proportions were regular.
Although neither contender gained a definite victory (for Berthollet was perfectly correct in his position over some of the difficult substances like glasses that he examined), by 1810, and in the light of Dalton’s theory, it was seen that the laws of definite and multiple proportions offered a securer foundation for quantitative chemistry. For the time being Berthollet’s views, which were eventually to illuminate physical chemistry and the theory of semiconductors, served only to confuse and handicap the development of analytical chemistry, which was so beguilingly explained by Dalton’s theory.
In 1808 William Hyde Wollaston and Thomas Thomson provided further convincing experimental examples of multiple proportions when they showed that there was a 2 : 1 ratio of CO2 in the bicarbonate and carbonate of potassium (viz. KHCO3 or K2O.2CO2.H2O and K2CO3 or K2O.CO2), and, for the same amount of acid in the normal and acid oxalates of potash and strontia, there was double the amount of base in the acid oxalates.
It was in the context of this experimental work with oxalates that Thomson recommended Dalton’s chemical atomism, having already briefly referred to it in the third edition of his important textbook, System of Chemistry, in 1807. Thomson’s initial account was based directly on conversations Thomson had had with Dalton in Manchester in 1804. Soon afterwards Dalton had read an account of his first list of atomic weights in a paper read to the Manchester Literary and Philosophical Society in October 1803 (published with differences in 1805). Thomson was also directly responsible for inviting Dalton to Scotland in 1807 to lecture on his views on air, heat and chemical synthesis to audiences at the Universities of Edinburgh and Glasgow. It was these lectures that led to the New System.
Although Dalton’s brilliant insight was developed by others, it is worth emphasizing that he retained other imaginative insights that remained undeveloped for decades. In particular, it is clear from surviving remnants that Dalton built models of atoms and compounds in order to illustrate his theory. This model-building followed directly from his first thoughts on mixed gases in 18034:
When an element A has an affinity for another B, I see no mechanical reason why it should not take as many atoms of B as are presented to it, and can possibly come into contact with it (which may probably be 12 in general) except so far as the repulsion of the atoms of B among themselves are more than a match for the attraction of an atom of A. Now this repulsion begins with 2 atoms of B to 1 of A, in which case the 2 atoms of B are diametrically opposed; it increases with 3 atoms of B to 1 of A, in which case the atoms of B are only 120° asunder; with 4 atoms of B it is still greater as the distance is then only 90°; and so on in proportion to the number of atoms. It is evident then from these positions, that, as far as powers of attraction and repulsion are concerned (and we know of no other in chemistry) binary compounds must first be formed in the ordinary course of things, then ternary, and so on, till the repulsion of the atoms of B (or A, whichever happens to be on the surface of the other), refuse to admit any more.
This statement shows that Dalton’s apparently intuitive appeal to a principle of simplicity in chemical synthesis was backed up by a geometrical force model – a model that, in a radically different setting, was to be used by ligand-field theorists a century and a half later. But it was entirely speculative, and, although it gave ‘order’ to Dalton’s symbols, it was not a path that the empirically minded Berzelius was to follow in his own symbolic language.
Dalton presented his theory within the context of ideas concerning heat at a time when the chemical world had become excited by the news of galvanic or current electricity. In 1800 the Italian physicist, Alessandro Volta (1774–1827), described his ‘voltaic
